Planning DV360 alongside paid social: roles, overlap and control

The question is rarely DV360 or social; it is what each should be for. Social platforms offer self-optimising delivery inside their walls. DV360 offers the rest of the internet - open-web display and video, YouTube with fuller controls, CTV, audio, DOOH - with deal-based buying and granular control that social abstracts away. A plan that treats them as interchangeable buys the overlap twice and the differences not at all.

Assign the roles before the budgets

Each role becomes its own plan line with a budget, KPI and flighting - reach and quality metrics for the DV360 awareness lines, conversion economics for social performance lines. Blending them into 'digital display and social' guarantees the wrong judgements later.

Plan the overlap, not around it

The same person scrolls a feed at lunch and reads the open web at night. Run social and DV360 prospecting against the same audience with no coordination and the combined frequency exceeds anything either dashboard shows - each platform caps within its own walls and knows nothing of the other.

• Set frequency intent per audience across the whole plan, then divide it: platform caps sized so the plausible combined exposure stays inside it.
• Use role separation as the natural de-duplicator: social on warm and lookalike audiences, DV360 on contextual, premium and CTV reach the feeds cannot buy.
• Exclude converters and active customers everywhere, and share suppression logic across both ecosystems.

Use the buying ladder as a control dial

DV360's real distinction is choice of buying mode. The open exchange buys scale at the lowest CPMs with verification treated as mandatory; PMPs buy known, premium environments at a quality premium; programmatic guaranteed buys certainty for tentpole moments at fixed prices. The plan should state the mode per line, because the mode is the quality and cost decision.

Deals need management: unmonitored PMPs habitually under-deliver while open exchange quietly backfills the impressions. A deal that is behind its committed volume is a pacing conversation with the publisher, not a mystery.

One plan, one spend definition, one taxonomy

Cross-ecosystem control fails on definitions before it fails on delivery. Agree whether budgets are gross or net including platform and data fees, align currencies and time zones, and apply one naming taxonomy so every DV360 insertion order and social campaign maps to its plan line automatically - the groundwork covered in cross-platform spend tracking.

Pace both sides against the same baseline

DV360 lines drift for their own reasons - deals under-delivering, floor prices above bids, creative stuck in QA, forecasted inventory that never materialised - and social drifts for its own. The discipline is identical: actual versus weighted planned-to-date per line, daily, with variance and required daily spend, per the pacing calculator.

Medusa holds social and DV360 lines in one approved plan, pulls actuals from the connected platforms - Google, Meta, TikTok, LinkedIn and DV360 - and flags drift across all of them in a single view, so reallocation weighs the whole allocation rather than one seat's dashboard.

Read results by role

Judged on last-click CPA, DV360 awareness lines will always lose to retargeting - that is the attribution model talking, not the media. Judge each line on its assigned job: incremental reach and quality delivery for the DV360 layers, conversion economics for the capture layers, and let incrementality thinking arbitrate where credit is contested.

Run the two ecosystems as one plan

Distinct roles, deliberate overlap management, one taxonomy and one pacing baseline - build the combined plan in Medusa and manage social and DV360 as a single allocation with daily control.